While measurements are useful, I'm still not sure if they are complete enough. Also, have had some experience with some of the speakers that have recently been mentioned.
Had a set of JBL M2 on loan here for several months. Thought they sounded awful, unlistenable for hi-fi use. One problem was the ugly sounding Class-D Crown amplifiers that JBL uses to get M2 published performance. Not very well designed for hi-fi, maybe good enough for PA use though. One other thing I didn't like was the bass porting. Bass sounded too flabby, LF transients were unrealistic. Tightly stuffing the ports to close them off left the speakers sounding over-damped. Ended up with the ports loosely stuffed with towels to where the LF transients were most listenable. Ditched the Crown amps, bi-amped instead with a Benchmark AHB2 for the horns, and a Bryston 4B for the woofers. Built an analog crossover with some opamps and the speakers starting sounding pretty good for box speakers. Of course FR was not good at that point since the Crown amps were programmed to do a lot of FR correction. Got some info and recommendations from a former JBL engineer who suggested designing passive crossovers, something he was working on for a similar speaker when JBL downsized and laid him off.
Another speaker I used long ago for hi-fi and for PA was Altec 604E. IIRC Big Red monitors used that same speaker with a custom horn exit aperture. Stock 604E had a multi-cellular arrangement that helped dispersion but sounded kind of distorted (all the old Altec multicellular horns sound distorted to me, never liked that approach nor did I like the sound of JBL acoustic lenses used for dispersion widening). For my home hi-fi 604E's, they were modified by cutting out the cells to make them about the same as the Big Red version. Sounded much better, less distorted that way.
The stories of the above two speakers are just to show that FR and dispersion are not the only things that are important. Sometimes they are not even the most important.
In a similar vein, the Yamaha NS-10 near field monitors were used to mix a lot of record releases. Last time I checked CLA was still using them to mix hit records for something like $10,000 per track. Point is FR is awful, but time domain transient response is especially good. Waterfall plots decay very fast and uniformly. Attached document goes into more detail. Mix engineers could learn how to reliably produce mixes that would 'translate' well to many different types of reproduction systems. Not the same end goal as for home hi-fi use.
Had a set of JBL M2 on loan here for several months. Thought they sounded awful, unlistenable for hi-fi use. One problem was the ugly sounding Class-D Crown amplifiers that JBL uses to get M2 published performance. Not very well designed for hi-fi, maybe good enough for PA use though. One other thing I didn't like was the bass porting. Bass sounded too flabby, LF transients were unrealistic. Tightly stuffing the ports to close them off left the speakers sounding over-damped. Ended up with the ports loosely stuffed with towels to where the LF transients were most listenable. Ditched the Crown amps, bi-amped instead with a Benchmark AHB2 for the horns, and a Bryston 4B for the woofers. Built an analog crossover with some opamps and the speakers starting sounding pretty good for box speakers. Of course FR was not good at that point since the Crown amps were programmed to do a lot of FR correction. Got some info and recommendations from a former JBL engineer who suggested designing passive crossovers, something he was working on for a similar speaker when JBL downsized and laid him off.
Another speaker I used long ago for hi-fi and for PA was Altec 604E. IIRC Big Red monitors used that same speaker with a custom horn exit aperture. Stock 604E had a multi-cellular arrangement that helped dispersion but sounded kind of distorted (all the old Altec multicellular horns sound distorted to me, never liked that approach nor did I like the sound of JBL acoustic lenses used for dispersion widening). For my home hi-fi 604E's, they were modified by cutting out the cells to make them about the same as the Big Red version. Sounded much better, less distorted that way.
The stories of the above two speakers are just to show that FR and dispersion are not the only things that are important. Sometimes they are not even the most important.
In a similar vein, the Yamaha NS-10 near field monitors were used to mix a lot of record releases. Last time I checked CLA was still using them to mix hit records for something like $10,000 per track. Point is FR is awful, but time domain transient response is especially good. Waterfall plots decay very fast and uniformly. Attached document goes into more detail. Mix engineers could learn how to reliably produce mixes that would 'translate' well to many different types of reproduction systems. Not the same end goal as for home hi-fi use.
I got very different experiences with the JBL M2, but not using the recommended crown, mainly because of the fan it has. It was for a rich friend, and we used Audiophonics build Purifi based amps and the then very new MiniDSP Flex with Dirac. And that is how i hear most used it, a high quality dsp with NCore or Purifi amps tuned custom by the installers of those (they are mostly used in post production studio's). With an analog crossover you can never make it work well.
JBL recoomend an amp from the same mother company (Harman/Samsung, that also owns Crown), that is logic from bussines perspective. But it's not because they do you should follow them blindly...
And the Altec multicell horn does not distort, it's the driver that distort if you feed it to much power. The whole driver should not get more than 50W and preferable even much lower. Most who complain about distortion use more. It's maybe not the best horn arround (certainly not) but on low power it suits the driver well, and with eq it sounds very good.
Actually, I didn't say they were distorted. Said they sounded distorted to me. I attributed it to multiple source interference. Each of the cells had some dispersion and or edge diffraction overlap with adjacent cells, so there appeared to be some frequency dependent interference. Again, to me it sounded like a type of distortion, where the word distortion is used in the general sense (not just HD/IMD).
EDIT: The interest in a analog crossover for the M2 speakers for to see if we could keep phono entirely in the analog domain. Once its digitized then run through a dac, it always sounds different. ADC->DAC is IME never perfectly transparent, most likely due to SD data conversion quirks producing artifacts other than HD/IMD.
Last edited:
This is all my opinion... But this thread is kinda what sucks the fun and life out of building speakers and even playing with audio. There is a joy in the journey, and regardless of what any arrogant know-it-all thinks, there is no perfect way to measure perfection in audio reproduction. The reality is... it's all fake, you aren't watching reality, you're listening to a fake production (even in a live performance, that's an amplified signal attempting to reproduce the actual performance). Not at all saying there is no science involved, there's a lot of science and (proven) methods that work in speaker building. And while I utilize those methods every time I build speakers... those methods will only get you so far, eventually you still need to be creative and use your brain. Anyone who says measurements are all you need... definitely could be building speakers that sound better if they go beyond just measurements. Use the measurements to find problems, use your ears and your soul to make them sound good.
Aside from that rant... Measurements are important in design, they help isolate problematic areas in the frequency response. Measurements are far from the end-all, be-all area of speaker design, it takes people to listen and decide if the speakers sound good, regardless of what the measurements say. I'm not really sure what the goal is here... would you even want a computer to spit out a perfect speaker?? Wouldn't that basically eliminate this whole forum? From there on we just 3d print our whole stereo, because "this computer says it's the way it should be"... I enjoy the journey, I enjoy building speakers, I enjoy the different nuances that separate one speaker from another and I'm glad no 2 are the same.
For the record, I wouldn't want a computer telling me what painting I need hanging in my living room either.
Aside from that rant... Measurements are important in design, they help isolate problematic areas in the frequency response. Measurements are far from the end-all, be-all area of speaker design, it takes people to listen and decide if the speakers sound good, regardless of what the measurements say. I'm not really sure what the goal is here... would you even want a computer to spit out a perfect speaker?? Wouldn't that basically eliminate this whole forum? From there on we just 3d print our whole stereo, because "this computer says it's the way it should be"... I enjoy the journey, I enjoy building speakers, I enjoy the different nuances that separate one speaker from another and I'm glad no 2 are the same.
For the record, I wouldn't want a computer telling me what painting I need hanging in my living room either.
While at the moment the technical part might not come easy to you, places like this forum are a great way to learn. IMO, the biggest thing that keeps people from having fun with this, is that there are so many people hollering about what is and isn't the right or wrong way to do it. Quit listening to those people, get yourself a decent amp / preamp or even a used receiver and give it a go. But be warned... it's an expensive hobby, but it's only expensive if you enjoy it 😉.
If you enjoy woodworking, and you enjoy music, you can build speakers.
I'd never heard of kii so I had a look.
This might be an example of a speaker that measures well (see: https://www.erinsaudiocorner.com/loudspeakers/kii_three/ even includes thd measurements) and sounds accurate.
As far as sounding "good" goes, as Lol Coxhill said: beauty is in the ear of the beholder
Edit: just read mattsk8's post and that's the triple truth, Ruth, right there.
Mastering is quite a different application from listening at home of course, but I'd still like to hear that $$$$kii
This might be an example of a speaker that measures well (see: https://www.erinsaudiocorner.com/loudspeakers/kii_three/ even includes thd measurements) and sounds accurate.
As far as sounding "good" goes, as Lol Coxhill said: beauty is in the ear of the beholder
Edit: just read mattsk8's post and that's the triple truth, Ruth, right there.
Mastering is quite a different application from listening at home of course, but I'd still like to hear that $$$$kii
Well, i personally also prefer some type of more coloured speakers. I'm in into single driver fullrange speakers. But i will never claim those are the most neutral speaker arround, they are not. But they sound good to my ears.
But on absolute neutral speaker, there are three brands that stand out in the market, Kii, Genelec and Neumann. All use a very different approach to get there but they measure the best at the end. And those also sound very good. But not as good to my ears as those single driver fullrange drivers speakers that i love.
I believe Hi-Fi does not exist, but there are high-quality pieces of Audio equipment.
When I listen to music I like to do it through a "traditional" high-quality system, albeit without exaggerations or fanatism because I like to reach the best value for money regarding every link in the chain in order to (try to) obtain the highest possible quality for my current budget.
I have not the attitude of exotic designs that anyway I could not "handle" as I would like and that however I also appreciate to know, making me fantasize for a while.
I've never built speakers (and I think I'll never be able to do it), but I like to optimize and slightly modify everything possible according to my ideas by evaluating results through listening until I achieve my psycho-acoustic satisfaction, that's my (temporary) pleasure.
Don't read too much into what I said, I believe hi-fi exists and I definitely believe some speakers / amps / preamps / etc... sound better than others. I just don't think we've addressed every measurement out there that could quantify all of the nuances that separate what someone might perceive as their favorite speaker from one that isn't their favorite speaker.
But also remember that in this hobby we have people who prefer tube amps over solid state amps, or solid state over digital... horn speakers over conventional, some people don't like ribbon tweeters, some people don't like dome tweeters and in dome tweeters some think they hate metal dome while others think they prefer metal dome... and on- and on- and on... we go. Just saying, none of those people are necessarily wrong, they might actually prefer one over the other. But doesn't mean someone with a different opinion is wrong, it just means I like Mexican food while you might prefer Italian.
Depends how you define hi-fi. Some people don't think a system requires absolute perfection to qualify as hi-fi. Maybe we could invent a new term: perfect-fi (which does not exist).
Different people have different beliefs about that. Sometimes I think of it this way (for example): do you want a measurement dac or a listening dac?
No DAC, I'm strictly analog, therefore, if you're using a DAC you're doing this wrong. (totally joking 😜)
Please, do not read too little in what I said...
Hi-Fi is only an acronym, but its interpretations are as many as the single interpretations of any Audio system owner (without mentionig the recordings).
To obtain "your" sonic preference shaping the choice of any link of an Audio chain with certain sound features put together for the purpose means going out of a pre-established concept, that de facto does not exist.
Maybe the measurements of every single device of every single system may be within a conventional standard range, but each system will fatally sound different from any other one.
Each system will have a different owner's sound imprint and I there are not two systems sounding in the same way.
So, (I believe) Hi-Fi does not exist.
Not to read literally, please. 😉
Last edited:
I love it all, I've enjoyed many different arenas in this hobby (digital, analog, tubes, solid state...). In my opinion, pigeon holing yourself into one specific arena of this hobby is akin to being racist- we might end up missing the most beautiful one in the world if we cling to dumb ideologies. Take the time to understand the science so you don't pay for useless snake oil, but mostly have fun with it.
I guess to circle back to the subject, measurements definitely matter if you're designing, but I'd never buy a pair of speakers solely based on measurements (and not actually hearing them) without considering that there's a strong possibility that I might not enjoy the speakers... Because measurements don't tell the whole story.
I guess to circle back to the subject, measurements definitely matter if you're designing, but I'd never buy a pair of speakers solely based on measurements (and not actually hearing them) without considering that there's a strong possibility that I might not enjoy the speakers... Because measurements don't tell the whole story.
Rest assured.
Agreed.
So much so that I've lately said/repeated it thousands of times all around. 🙂